Large Scale Online Learning of Image Similarity Through Ranking

By Gal Chechik, Varun Sharma, Uri Shalit, Samy Bengio. 2009

917 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Representation & Self-Supervised Learning

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

In Representation & Self-Supervised Learning

Citations in the corpus

Revisiting IM2GPS in the Deep Learning Era (2017, neutral): “[5, 19, 35].”
Dual Contradistinctive Generative Autoencoder (2020, neutral): “Contrastive learning can be connected to metric learning [4, 9, 5].”
DeltaPhish: Detecting Phishing Webpages in Compromised Websites (2017, background): “This is a well-known practice in the area of machine learning, usually referred to as similarity learning [40].”
Understanding How Dimension Reduction Tools Work: An Empirical Approach to Deciphering t-SNE, UMAP, TriMAP, and PaCMAP for Data Visualization (2020, methodology): “Nevertheless, techniques such as triplets are used for both embedding (Chechik et al., 2010; Norouzi et al., 2012) and DR (Hadsell et al.”
Deep Metric Learning via Lifted Structured Feature Embedding (2015, background): “This function poses two computational challenges: (1) it is non-smooth, and (2) both evaluating it and computing the subgradient requires mining all pairs of examples several times.”
Supervised Contrastive Learning (2020, background): “ch have achieved excellent performance in self-supervised learning in recent years in the image and video domains [50,25,21,19,46,6,43] and have connections to the large literature on metric learning [48,5]. As the name suggests, contrastive losses consist of two “opposing forces”: for a given ancho”
Learning Fine-Grained Image Similarity with Deep Ranking (2014, methodology): “A comparison of the ranking examples of ConvNet, OA-SIS feature (L1HashKPCA features with OASIS learning) and Deep Ranking is shown in Fig.”
Deep Metric Learning Using Triplet Network (2014, background): “We follow a similar task to that of Chechik et al. (2010).”
Bayesian representation learning with oracle constraints (2015, background): “In Chechik et al. (2010), a metric respecting the particular distances in similarity is learned.”
Poisoning and Backdooring Contrastive Learning (2021, methodology): “Early techniques would train using a triplet loss (Weinberger & Saul, 2009; Chechik et al., 2010) to distinguish two similar objects from a third different object.”
A Survey on Metric Learning for Feature Vectors and Structured Data (2013, background): “OASIS31 (Chechik et al., 2009, 2010) learns a bilinear similarity with a focus on large-scale problems.”
Hamming Distance Metric Learning (2012, methodology): “Such loss functions have been used in metric learning [5, 11], and, as shown below, they are also naturally suited to Hamming distance metric learning.”
ProxyNCA++: Revisiting and Revitalizing Proxy Neighborhood Component Analysis (2020, background): “image and an image similar to anchor image) and negative pair (an anchor image and an image dissimilar to anchor image) [3].”
VSE++: Improving Visual-Semantic Embeddings with Hard Negatives (2017, methodology): “Inspired by the use of a triplet loss for image retrieval (e.g., [4, 8]), recent approaches to joint visual-semantic embeddings have used a hinge-based triplet ranking loss [13, 15, 29, 36]: where α serves as a margin parameter, and [ x ] + ≡ max ( x , 0 ) .”
Privacy Regularization: Joint Privacy-Utility Optimization in LanguageModels (2021, neutral): “…use the language model loss ( L LM ) of the previous section (Eq 1), and we set the privacy loss to be the triplet loss: The triplet loss is commonly used in vision tasks for training embeddings that map images from same category to neighboring points in the embedding space (Chechik et al., 2010).”
Adapting Visual Category Models to New Domains (2010, background): “Finally, metric and similarity learning has been successfully applied to a variety of problems in vision and other domains (see [6,11,14,5] for some vision examples) but to our knowledge has not been used for domain adaptation.”
What comprises a good talking-head video generation?: A Survey and Benchmark (2020, methodology): “It is worth mentioning that we also tried to use metric-learning to refine the lipreading network with ArcLoss and Triplet loss (Chechik et al. 2010).”
Deep Metric Learning with Angular Loss (2017, background): “Despite the various forms, the major work of deep metric learning can be categorized as minimizing either the contrastive loss ( a.k.a. , Siamese network) [6] or the triplet loss [34, 5].”
PCCA: A new approach for distance learning from sparse pairwise constraints (2012, background): “In contrast, On-line Algorithm for Scalable Image Similarity (OASIS) [5] and Probabilistic Relative Distance Comparison (PRDC) [33] are specifically designed to work with pairwise constraints.”
Deep Learning for Content-Based Image Retrieval: A Comprehensive Study (2014, methodology): “Following exactly the same experimental setting in [7], we use subsets of sizes 10, 20, 50 classes, and images from each class were randomly split into a training set of 40 images and a test set of 25 image.”